- fix rounding bug in
cost_knapsack
for better integer approximations
- add documentation in all
R
source files for estimation functionality:-
cate.R
-
data.R
-
effects.R
-
fluctuation.R
-
nuisance.R
-
costs.R
-
segment.R
-
confint.R
-
setup_data.R
-
wrappers.R
-
Lrnr_prob_known.R
-
- add documentation where necessary in all
R
auxiliary source files:-
bound.R
-
scale.R
-
globals.R
-
options.R
-
zzz.R
-
- fix TMLE fluctuation to fit each parameter separately
- add confidence intervals for CATE and effect measure estimates
- add in checks for diagnostics at smaller sample sizes
- reduce to a few high-level functions:
sherlock
to detect segments,watson
to assign an optimal rule, andmycroft
to compute effects
- rename HTE to OTE, since HTE already defined otherwise in ABlaze
- evaluate OTE conditional on segmentation covariates, ie, COTE = E(OTE|V)
-[x] create a wrapper function to bundle set_est_data
and est_cate
-[x] add summarization of EIF across IDs for repeated measures
-[x] add functionality to handle known g in PS estimation
-[x] export global option for sl3.pcontinuous
to avoid guessing categorical